Revisiting dissociation hypotheses with a structural fit approach: The case of the prepared reflex framework

Attitude and social cognition research often tests dissociations in performance on “explicit” and “implicit” measures using tasks that widely differ from each other. This prevents a clear interpretation of the findings. A structural fit approach, involving tasks that differ only on the factor of theoretical interest, should be preferred. Here, we revisited a dissociation hypothesis using a structural fit approach for the first time. Specifically, Melnikoff, Lambert, and Bargh's (2020) Prepared Reflex framework posits that the establishment and activation of an association between a planned action and its target make the valence of the action unintentionally spread to the target. As a result, the target is evaluated according to the valence of the action. Once the action plan is inactive, however, unintentional measures of attitudes should no longer reflect the valence of the planned action, as the association is not active anymore. In contrast, the valence of the inactive plan should continue to influence intentional measures of attitudes. Melnikoff et al. (2020) found support for this dissociation hypothesis in studies comparing evaluations on self-reports versus an Affect Misattribution Procedure (AMP): Whereas both measures captured the valence of the planned action when the plan was active, only the self-reports captured this valence after the plan was inactive. Here, contrary to the dissociation hypothesis, the plan valence influenced target evaluations on an intentional AMP when the plan was active but not when it was made inactive. These findings highlight the importance of using structurally-fitted tasks in attitude and social cognition research.